pKa and logP for caffeine

The predicted pKa and logP values for caffeine differ quite a bit from literature values.





pKa = 1.5 (marvin) vs 10.4 (SRC)





logP = -0.79 (marvin) vs -0.07 (SRC)





Any ideas?

Dear Carl,








I think the large value (10...) is not pKa. It must be pKb.





I have obtained -0.54 for the logP with the new version.(will be released soon)








Jozsi